Microcellular Foams - American Chemical Society

Chapter 7

Microcellular Foams

Downloaded by NORTH CAROLINA STATE UNIV on August 20, 2012 | http://pubs.acs.org Publication Date: June 1, 1997 | doi: 10.1021/bk-1997-0669.ch007

Vipin Kumar and John E . Weller UW-Industry Cellular Composites Consortium, Department of Mechanical Engineering, Box 352600, University of Washington, Seattle, W A 98195

Microcellular polymers refer to closed cell thermoplastic foams with an extremely large number of very small cells. Typically, one hundred mil lion or more bubbles nucleate per cm of polymer, and each cell is of order 10 μm diameter. These materials were first produced in the early 1980's with the objective of reducing the amount of polymer used in mass produced items, and have the potential to revolutionize the way thermo plastic polymers are used today. Microcellular plastics have been pro duced from a number of polymers and can be produced with densities ranging from 3 to 99 percent of the density of the solid material, offering engineers a new range of properties for design. Microcellular foams pos sess improved fatigue life and energy absorption characteristics as well as a high specific strength. It is expected that these materials will be utilized in a wide variety of applications in the next century. 3

Imagine solid polymers replaced by microcellular polymers with 20 to 40 percent reduc tion in material without compromising their function or required strength. Also look forward to composite microcellular materials with layers of solid and cellular materials designed to offer optimum mechanical and thermal properties and a minimal use of poly mer. Created by processes without using any chemicals that are harmful to the environ ment, such materials may well become widely used by the turn of the century! Microcellular polymers are closed cell plastic foams with a bubble densities in excess of 100 million bubbles per cm and diameters of order 10 μπι. The idea to intro duce very small bubbles in plastics by gas nucleation was originally advanced by Profes sor Nam P. Suh ( i ) of the Massachusetts Institute of Technology around 1980 as a means to reduce the cost of many mass produced plastic items. The rationale was that if bubbles smaller than the critical flaws that already exist in plastics could be introduced in suffi cient numbers, then the material density could be reduced while maintaining the essen tial mechanical properties. The basic microcellular process is a two-step process shown in Figure 1. In the first step, the polymer sample is saturated by a pressurized, non-reacting gas in a pres sure vessel. Over time, the polymer absorbs the gas, and eventually become saturated. When the polymer is saturated, it will not absorb any more gas, and the gas is distributed uniformly throughout the polymer. Typically, it takes several hours to several days for a thin polymer sheet to become saturated. Of course, the amount of time required to satu3

© 1997 American Chemical Society

In Polymeric Foams; Khemani, K.; ACS Symposium Series; American Chemical Society: Washington, DC, 1997.

101


102

POLYMERIC FOAMS

Figure 1. Schematic of the microcellular process: (a) saturation of polymer with gas at room temperature; (b) bubble nucleation and growth at atmospheric pressure.



7.

KUMAR & WELLER

Microcellular Foams

103

rate the polymer depends on the polymer itself, its diffusivity, its thickness, the satura tion temperature, and sometimes the gas pressure. This saturation step is usually carried out at room temperature since the solubil ity of most gases is higher at lower temperatures. Higher saturation temperatures may be employed however, to accelerate the diffusion of gas in the polymer. After the polymer has become saturated, it is removed from the pressure vessel and then foamed. In the second step of the process, the saturated sample is heated to the foaming temperature, which is close to the glass transition temperature of the virgin polymer. During this stage a large number of bubbles nucleate and grow. Bubble growth is controlled by holding the specimen at the foaming temperature for a certain length time. This time is commonly referred to as the foaming time. Since the growth rates of bubbles near the glass transi tion temperature are relatively low, it is possible to arrest the bubble growth in a control lable manner by quenching the foam in a low temperature bath. Figure 2 shows scanning electron micrographs of a number of polymers foamed by this process. The microcellular structure for all of these foams is very homogenous, and the bubbles are typically between 5 to 25 μπι in diameter. One exception is the PET foam, where the bubbles are an order of magnitude larger. The average bubble size in all microcellular foams is strongly influenced by the number of bubbles that nucleate, which in turn is a function of the concentration of gas in the sample after saturation. The den sity of the foams can be controlled by choosing an appropriate gas saturation pressure, and using a suitable foaming time and temperature. These parameters are specific to a given gas-polymer system, and for the most part must be determined experimentally. Figure 3 shows scanning a scanning electron micrograph of a polystyrene sample showing the full specimen cross section. This sample was originally 0.40 mm thick and grew to a thickness of 0.65 mm after foaming. This example shows that thin sections of polymers normally made from solid polymer can indeed be foamed by the microcellular process (7 ). Overview of Microcellular Polymers The basic microcellular process outlined above has been successful with a number of amorphous polymers, and semicrystalline polymers with low levels of crystallinity. Some of these gas-polymers systems are discussed below. Polyvinyl Chloride (PVC). Microcellular PVC foams have been produced by Kumar and Weller (2 ) using carbon dioxide as the nucleating gas. Bubble densities in the range 10 -10 per cm of polymer were achieved, and foams with relative density (density of the foam divided by the density of the polymer) in the range 0.15 to 0.94 have been produced. The sorption data for carbon dioxide in PVC (2 ) shows that for 2 mm thick specimens of a rigid PVC formulation saturated at 4.8 MPa and room temperature, the gas concentration at equilibrium is approximately 75 mg of carbon dioxide per gram of PVC, or about 7.5 percent by weight. Thus a very high concentration of gas molecules exists in the polymer matrix at saturation. As a result, there is significant plasticization of the polymer, and the glass transition temperature drops from 78.2 °C for original P V C to an estimated 14 °C upon saturation (2 ). Bubble nucleation and growth is therefore possible at very moderate temperatures, at or below the glass transition temperature of the original polymer. In the case of PVC, the lowest temperature at which microcellular foam was produced was 14 °C Figure 4 shows the range of relative densities for the family of microcellular P V C foams produced by heating the saturated PVC specimens to different temperatures. The lowest density foams (15 to 20 percent of the solid PVC density) are produced in the 100-115°C temperature range. Note the nearly linear variation of foam relative density with foaming temperature in the range 56°C to 105°C, demonstrating that the foam den sity can be simply and accurately controlled by controlling the temperature to which the saturated plastic is heated. 7

9

3



104

POLYMERIC FOAMS

Figure 2. Scanning electron micrographs showing examples of the structure created by the microcellular process: (a) PVC; (b) polycarbonate; (c) ABS; and (d) PET.

Figure 3. Overall view across a polystyrene sample originally 0.40 mm thick (7). The thickness after foaming was 0.65 mm.


7.

105

Microcellular Foams

KUMAR & WELLER

Polycarbonate (PC). Creation of microcellular structure in polycarbonate was first re ported in 1990 by Kumar, Weller and Hoffer (3 ). Carbon dioxide, due to its high solubil ity in polycarbonate, was used to achieve very high bubble nucleation densities in the range 1-10 χ 10 per cm of polymer. Sorption data shows an equilibrium gas concentra tion of approximately nine percent by weight at a saturation pressure of 4.8 MPa. The number of bubbles nucleated was found to increase exponentially with increasing gas saturation pressure. A family of microcellular polycarbonate foams has been produced ranging in density from 10 to 99 percent of the density of solid polycarbonate. Downloaded by NORTH CAROLINA STATE UNIV on August 20, 2012 | http://pubs.acs.org Publication Date: June 1, 1997 | doi: 10.1021/bk-1997-0669.ch007

9

3

Weller (24) has conducted a detailed characterization of the P C - C 0 system. The den sity of microcellular polycarbonate foams as a function of the foaming temperature for various saturation pressures is shown in Figure 5. Again, we see a linear relationship, suggesting that the foaming temperature is a critical parameter for controlling the density of the microcellular foams, and thus an effective means of in-process control of foam morphology. The data in Figure 5 is discussed further in the section on mechanical properties. 2

Polystyrene (PS). Microcellular polystyrene has been studied by a number of investiga tors (1,6,7,8,9,10,15 ). Martini et al. (1 ) produced microcellular polystyrene with cells in the 5-25 mm range and void fractions of 5 to 30 percent. Using higher nitrogen pres sures for saturation, Kumar (10 ) extended the void fraction range to 80 percent. This was possible because the number of bubbles increased exponentially with saturation pres sure, allowing for lower density foams to be produced while keeping the average bubble diameter in the microcellular range. Polyethylene Terephthalate (PET). Kumar and Gebizlioglu (11 ), and Baldwin and Suh (12 ) have applied the microcellular process to foam PET using carbon dioxide. The results with PET have been most interesting. It has been found that with sufficiently long exposure to high pressure carbon dioxide, PET begins to crystallize. The resulting PET foams have a significantly higher crystalline content, and although mechanical property data is not yet available, these foams appear to be much stronger than microcellular foams from other polymers at comparable relative densities. P E T G . The most recent addition to the family of polymers successfully processed by the microcellular process is PETG (22,23). Relative densities of three percent and higher have been achieved using carbon dioxide for bubble nucleation, see Figure 6. An inter esting observation in C 0 - P E T G system is the relatively high foaming temperature to which the process appears to be stable. In Figure 6, the highest temperature at which foam was produced is 120 °C, which is nearly 40 °C above the Τ of PETG. By compari son, in Polycarbonate-C0 system, the process becomes unstable at 160 °C, about 10 °C above the T of polycarbonate. 2

2

g

Mechanical Properties of Microcellular Polymers The available mechanical property data on microcellular foams is expected to increase as the number of researchers and our understanding of the processing characteristics of different gas-polymer systems grows. However, the data available so far is quite limited except for a few select systems. The results to date show that these novel materials offer many properties that were not previously available from either conventional foams or solid polymers. In many cases, the properties are improved over those of conventional foams, and in some cases a particular property may be better than that of the original, solid material. The polycarbonate - C O system has received a great deal of attention since a wide range of processing conditions have been documented (24). Figure 5 shows the ?



106

POLYMERIC FOAMS

50

60

70

80

90

100

110

120

130

Foaming Temperature (°C) Figure 4. Plot of PVC foam relative density as a function of foaming temperature for specimens saturated with carbon dioxide at 4.8 MPa ( 2)



7.

KUMAR

& WELLER

Microcellular Foams

107

steady state process space for PC-CO system, where steady state refers to the state where all cell growth has completed. This figure illustrates the wide range of relative densities that can be produced in this system and the numerous processing conditions that can be employed. These different processing conditions can be used to produce foams that have the same relative density and different average cell size. For example, a sample could be saturated at 5 MPa and foamed at approximately 100 °C or it could be saturated at 1 MPa and foamed at 140 °C to produce a foam of the same density. The possibility of produc ing these foams is illustrated by the horizontal dashed line in Figure 5. For the two processing conditions discussed above, a relative density of approximately 0.5 will be produced. In addition, foams of different average cell size are produced as illustrated by the two indicated data points, which have the same relative density and cell sizes of approximately 5 μτη and 22 μηι. The ability to produce foams of the same density and different cell size has opened the possibility to explore the effect of cell size on the properties of microcellular materials. The strength of microcellular foams is nearly proportional to the foam density. Figure 7 shows the relative tensile strength of several microcellular foams and some structural foams as a function of the relative density. For this plot, the relative tensile strength was computed by dividing the strength of the foam by the strength of the un processed material. The relative tensile strength of microcellular foams appears to be marginally better than the strength of conventional foams. It is interesting to note that microcellular foams of relatively small thicknesses retain strengths similar to structural foams which have far thicker cross-sections. When compared to the rule of mixtures line, the data for all microcellular foams is slightly lower than this prediction. A l l of the microcellular data is significantly higher however, than the line representing the com monly used relative density squared model (25). The tensile modulus of several microcellular foams is shown in Figure 8 . Here results are shown from Weller (24) who tested microcellular PC, microcellular ABS, and microcellular PET; Kumar and VanderWel (14), where microcellular PC was tested; and Waldman (15) who tested microcellular PS. A l l of the relative modulus values shown in this figure were computed by dividing the modulus of the foam by the modulus of the best estimate of the matrix material. For the amorphous polymers, PC, ABS, and PS, the modulus of the matrix material was taken as the modulus of the original, unprocessed material. In the case of PET the properties were taken as those of the C 0 crystallized material (see (24)). The error bars in this figure represent the relative standard deviation. Figure 8 also shows a plot of the Mori-Tanaka model, adapted for foams (24) 2

fPf] [PmJ

w4l fPf]

p |

(1)

f

PmJ

E = modulus of the foam (MPa) E = modulus of the matrix material (MPa) Pf = density of the foam (g/cm ) ρ = density of the matrix material For all of the microcellular systems, the Mori-Tanaka model can be used to predict the relative tensile modulus over a wide range of relative densities. Waldman (15) tested the toughness of several microcellular polystyrene foams and found that the toughness increases nearly 400 % as a result of the enhanced craze f

m

3

m



108

POLYMERIC FOAMS

140

Foaming Temperature (°C) Figure 6. Relative density as a function of foaming temperature for PETG saturated at 4.8 MPa (700 psi).

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Relative Density Figure 7. Plot of relative tensile strength of microcellular foams and some structural foams as a function of relative density (24). In Polymeric Foams; Khemani, K.; ACS Symposium Series; American Chemical Society: Washington, DC, 1997.


7.

KUMAR & WELLER

Microcellular Foams

109

initiation by the voids. Collias and Baird (26) repeated Waldman's experiments and performed several others, trying to show that the microcellular structure actually increased the toughness of the material. They concluded that the toughness of microcellular polystyrene only increases marginally over the original material. They also found that the presence of nitrogen gas in the polymer significantly affects the tensile toughness; the toughness was improved in the presence of nitrogen, and then decayed as the gas diffused from the polymer. In 1993, Wing and Kumar presented work on the creep response of polycarbonate and microcellular polycarbonate (27), the first time-dependent study on microcellular foams. They found that C 0 residing in the matrix can have a significant effect on the creep/creep recovery of microcellular foams. In addition, even after all of the measurable gas had left the matrix, it was found that the creep response continued to change as the specimen aged. Changes in behavior were observed for periods up to 60 days after all of the gas had diffused from the matrix, 90 days following removal of the material from the pressure vessel. Schapery's theory was applied successfully to characterize the time dependent response of both PC and microcellular PC. Seeler and Kumar (16) have performed extensive studies regarding the fatigue life of microcellular polycarbonate foams. Stemming from the original rationale for developing microcellular foams, it has long been hypothesized that properties like fatigue life should be significantly improved in microcellular foams due to crack blunting by the voids. Fatigue data from some of their experiments are shown in Figure 9. The unfoamed material in this figure represents tests performed on unprocessed, virgin polycarbonate. Data at other relative densities are microcellular PC foams, and show similar behavior to the solid material. Specimens with a relative density of 0.90 (approx. 10% reduction in density) show the same fatigue life as the solid polymer for nominal stresses between 25 and 45 MPa. The fatigue life of the 0.97 relative density foam exceeds that of the solid polymer by an order of magnitude at a nominal stress of approximately 30 MPa. Thus the introduction of a microcellular structure by the solid-state process appears to increase the fatigue life of very high relative density polycarbonate foams. This phenomenon should be investigated in other gas-polymer systems. 2

Innovations in Microcellular Processing Initially it was believed that the bubble nucleation in the microcellular process is caused by the supersaturation of the polymer (I, 6 ). Kumar (7) showed that the reduction in the solubility of the gas with temperature can provide a very significant driving force for bubble nucleation even without creating supersaturation condition. Thus an alternative to the process shown in Figure 1 would be to saturate the polymer at a certain temperature and pressure (this could be room temperature or an elevated temperature) and then raise the temperature of the polymer to cause bubble nucleation while maintaining the gas pressure at the level used for saturation. This principle has been employed by Park and Suh (17) to create a microcellular structure in extruded filament. This advance is a significant step toward production of microcellular polymers in a continuous process. A thermoforming process to produce microcellular parts was developed by Kumar (7, 18). In this process a polymer sheet saturated by gas is thermoformed using a heated mold. Thus the shape of the part is achieved first, and the foaming is completed later. This process can be readily adapted to produce microcellular components by thermoforming from a plastic sheet. Figure 10 shows a picture of a microcellular polystyrene container produced by this process. A skin of unfoamed polymer can be created on the microcellular foam by letting the gas diffuse out of the surface layers prior to foaming (19 ). This simple modification to the original process is illustrated in Figure 11. In the region where the gas concentration in the polymer drops below the minimum concentration needed for nucleation C*, bubbles do not nucleate when the polymer is heated resulting in an unfoamed skin. The



110

POLYMERIC FOAMS

0.2 L0.4

>

I 0.5

ι—

I 0.6

ι

I 0.7

.

I 0.8

•

I 0.9

ι

I 1.0

Relative Density ( p/P ) m

Figure 8. Plot of relative tensile modulus of several microcellular foams as a function of the relative density. The solid line is the Mori-Tanaka prediction (Eqn. I) (24).

ι—ι—I

ο

H

50.0

1

ο

ο

60.0

Î I III

ο

I

ι I ι ι ιι

r~

unfoamed relative density = 0.97 relative density = 0.90

•

ο ο ο

I

•

ο 8

°

A

• Ο" Ο

40.0

* A

30.0 ο • ο ο 20.0

I

10

J

10

4

10

5

10

ι

ι

b

ι I I I I

10

7

Life (Cycles) Figure 9. Fatigue life of high relative density microcellular polycarbonate foams (16). In Polymeric Foams; Khemani, K.; ACS Symposium Series; American Chemical Society: Washington, DC, 1997.

KUMAR & WELLER

Microcellular Foams


7.

Figure 10. Example of a microcellular part made by a modified thermoforming process: (a) photograph of a microcellular container; (b) scanning electron micrograph of a sample from container bottom; (c) from the container wall. The average cell size is approximately 7 μτη (18).


111


112

POLYMERIC FOAMS

integral skin thickness at time t

foamed core Figure 11. Schematic showing development of a skin region as a function of gas desorption time. C* is the minimum concentration needed for bubble nucleation (19).

0» Figure 12. Illustration of control of microstructure: (a) schematic of foamed core with solid skin produced by the process shown in Figure 11 ; (b) schematic of an inverted structure with a foamed skin and a solid core (21). In Polymeric Foams; Khemani, K.; ACS Symposium Series; American Chemical Society: Washington, DC, 1997.


7.

KUMAR & WELLER

Microcellular Foams

113

thickness of this solid skin can be easily controlled by monitoring the time allowed for gas desorbtion and the foaming temperature. Thus microcellular foams with an engi neered skin/core structure can be created for applications where the solid material or foamed material alone cannot be used. In PET foams a crystalline skin integral to the foam core can be produced (11 ) creating a cellular composite material with higher spe cific strength. Research on nondestructive evaluation of the microcellular structure, and on de velopment of sensors for in-process control of the microstructure has been initiated (20, 21 ). The controllability of microstructure is illustrated in Figure 12. These structures have been realized by Holl (21) by control of the gas concentration profile in the speci men. These advances point to the possible creation of composite structures designed to deliver engineered mechanical and thermal properties with minimal use of polymeric materials. Impact of Microcellular Polymers There is a recognized need for minimizing the consumption of our natural resources. Further, there is an increased awareness that in processing our natural resources, the environment must be protected. Microcellular processes and materials address both of these concerns, and are therefore likely to strongly impact the way plastics are used in the future. Wherever solid plastics are used today, there is the potential of significant reduc tion in the amount of material used by employing microcellular plastics. The savings in material and transportation costs could potentially be very significant for a number of plastic items. In aircraft, for example, the weight saving by use of microcellular plastics could significantly enhance payload efficiency. Although the mechanical property data available so far is limited, there is an indication that materials produced by this process could show improved toughness and fatigue life over solid plastics. In the future, it is expected that these materials will be used in a number of applications (e.g. packaging, automotive, home construction, toys, aerospace, medical prostheses, telecommunications, to name a few). As no harmful chemicals are used in the microcellular process, it is likely that microcellular polymers will replace many cellular polymers now produced by processes that damage the environment. The ability to control the microstructure and create composite structures will greatly increase the range of applications of these mate rials. We expect that by the turn of the century, the engineering and commercial potential of these novel materials will begin to be realized. Acknowledgments Research in microcellular polymers at the University of Washington has been funded by the National Science Foundation Grant MSS 9114840, the Washington Technology Cen ter, and by the University of Washington — Industry Cellular Composites Consortium. This support is gratefully acknowledged. References

1. Martini J. E.; Waldman F. Α.; Suh Ν. P.; SPE Technical Papers 1982, Vol. 28, pp 674-676. 2. Kumar V.; Weller J.E International Polymer Processing 1993, Vol. 13, pp 73-80. 3. Kumar V.; Weller J. E.; Hoffer Η. Y. Processing of Polymers and Polymeric Com posites, Vol. MD-19, ASME: New York, 1990, p 197-212. 4. Durril P.L.; Griskey R.G. AIChE Journal, 1969, Vol. 15, pp 106-110. 5. Kumar V.; Weller J. E., Journal of Engineering for Industry, 1993, Vol. 116, pp 413-421.



114

POLYMERIC FOAMS

6. Colton, J. S. The Nucleation of Thermoplastic Microcellular Foam, Ph.D. Thesis, Mechanical Engineering., MIT, 1985. 7. Kumar V. Process Synthesis for Manufacturing Microcellular Thermoplastic Parts: A Case Study in Axiomatic Design, Ph.D. Thesis, Mechanical Engineering, MIT, 1988. 8. Ramesh N. S.; Kweeder J. Α.; Rasmussen D. H.; Campbell G. A. SPE Technical Papers, 1992, Vol. 38, pp 1078-1081. 9. Kweeder J. Α.; Ramesh N. S.; Campbell G. Α.; Rasmussen, D. H. SPE Technical Papers, 1991, Vol. 37, pp 1398-1400. 10. Kumar V.; Suh N. P. SPE Technical Papers, 1988, Vol. 34, p 715. 11. Kumar V.; Gebizlioglu O. S. SPE Technical Papers, 1991, Vol. 37, pp 1297-1299. 12. Baldwin D. F.; Suh N. P. SPE Technical Papers, 1991, Vol. 37, p 1300. 13. Vanderwel M., Microcellular Polycarbonate: Characterization of Microstructure and Mechanical Properties, M.S. Thesis, University of Washington, 1991. 14. Kumar V.; Weller J. E.; VanderWel, M.; Seeler K. A. Journal of Engineering Materials and Technology, 1993, Vol. 116, pp 439-445. 15. Waldman, F.A., The Processing of Microcellular Foam, S.M. Thesis, Mechanical Engineering MIT, 1982. 16. Kumar V.; Seeler K. A. Journal of Reinforced Plastics and Composites, 1993, Vol. 12, pp 359-376. 17. Park C. B.; Suh N. P., In Cellular Polymers; Editors, V. Kumar; S.G. Advani; MD Vol. 38, ASME, New York, 1992; pp 69-91. 18. Kumar V.; Suh N. P., Polymer Engineering and Science, 1990, Vol. 30, pp 13231329. 19. Kumar V.; Weller J. Ε SPE Technical Papers, 1992, Vol. 38, pp 1508-1512. 20. Garbini J. L.; Holl M.; Kumar V.; Murray W. M.; Jorgensen J. E. SPE Technical Papers, 1992, Vol. 38, p 1519. 21. Holl M . Dynamic Analysis, Measurement, and Control of Cell Growth in Solid State Polymeric Foams, Ph.D. Thesis, University of Washington, Seattle, 1995. 22. Kumar, V.; Eddy, S.; Murray, R. SPE Technical Papers, 1996, Vol. 42, pp 19201925. 23. Kumar, V.; Eddy, S.; Murray, R. Polymer Preprints, American Chemical Society, 1996, Vol.37, pp 779-780. 24. Weller, J. E. The Effects of Processing and Microstructure on the Tensile Behavior of Microcellular Foams, Ph. D. Thesis, University of Washington, Seattle, 1996. 25. More, D. R.; Couzens, Κ. H.; and Iremonger, M. J Journal of Cellular Plastics, 1974, Vol. 10, pp 135-139. 26. Collias, D. I.; Baird, D. G. Polymer Engineering and Science, 1995, Vol. 34, pp 1167-1177. 27. Wing, G. T.; Pasricha, Α.; Tuttle, M.; Kumar, V Polymer Engineering and Science, 1995, Vol. 35, pp 663-670. 28. Wendle, B. C. Engineering Guide to Structural Foam, Technomic, Westport, CT, pp 92-97.


Microcellular Foams - American Chemical Society

Recommend Documents